Objectives

By the end of this session, you should be able to:

• State the benefits of a reproducible analysis
• Login to your RStudio instance and explain the layout
• Illustrate the use of the Markdown language
• Explain how to seek help and create a reproducible example

Reproducibility

Workflows

Simplified Model

Tools

• RStudio: provides a nicer interface for using R and an authoring framework for data science
• LaTeX: helps behind the scenes to knit to PDF and offers more control over typesetting
• Zotero, BibDesk, and .bib files: manage references

Using Duke’s Docker-ized version of RStudio Server

In this workshop, we will use RStudio via the web.

• Login to your instance by going to https://vm-manage.oit.duke.edu/containers and entering your NetID.
• Click on Docker
• Click on RStudio

RStudio Interface

RStudio Interface

Projects

Never hardcode a working directory again!

• Click on Projects
• Choose to create a new project
• Follow the prompt to create a new directory called “ieat”

Organization

Organization

New files

R script

Create a new R script file

• similar to Stata’s do file
• not interactive
• output appears in console and plot window

RMarkdown

Create a new RMarkdown HTML file

Markdown language

  Plain text
  *italics*
  **bold**
  `monospaced font`
  ~~strikethrough~~ 
  sub/superscript^2^~2~
  endash: --, emdash: ---
  equation: $A = \pi*r^{2}$
  $$E = mc^{2}$$
  > block quote 
  list:
  * item 1
  * item 2
  * item 3

Getting help

• Google
• Read the documentation
• Stack Overflow
• Github issues / listserves for specific packages
• New: https://community.rstudio.com/

An unexpected result!

x <- c(4, 2, 6, 9, 2, 5, 4, 3, 7, 6, 8, 4, 7, 2, NA, 3)
x_mean <- mean(x)
x_mean

## [1] NA

?function

Stack Overflow

Bad questions

Reproducible Example